在「Quiz: Can we guess your age and income, based solely on the apps on your phone?」這篇給了一個測驗，透過回答 32 題 Yes/No 的答案來猜測你的性別、婚姻狀態與收入範圍，大約是 61% 到 82% 的準確率：(為什麼會是一個範圍？)
Based on those models, they then found that they could predict a user’s gender, age, marital status and income with between 61- and 82-percent accuracy.
會猜測四個答案，所以是 16 種組合：
There are 16 possible results, based on your gender (male/female), your age (over/under 32), your marital status (married/single) and your income (over/under $52,000).
我猜測用的方式與很久前 pest 講的「網路廣告商怎麼知道你是誰? 從 ClickStream 來判斷用戶資料」的方法類似，用已知的資料去 train 出一個模型，再丟進去判斷...
衛報報導了從 GitHub 上分析 pull request 的性別分析研究：「Women considered better coders – but only if they hide their gender」，原始論文出自「Gender bias in open source: Pull request acceptance of women versus men」。
研究的結果說明女性的 pull request 接受機率比男性高，但如果貢獻者可被確認是女性的話則會反過來，也就是說男女歧視問題是可被觀察到的：
Surprisingly, our results show that women's contributions tend to be accepted more often than men's. However, when a woman's gender is identifiable, they are rejected more often. Our results suggest that although women on GitHub may be more competent overall, bias against them exists nonetheless.
由於性別資訊不是必填項目，論文裡面也有提到透過 social network 的資料比對，以及其他方式去推測。這個研究成果看起來應該會產生不少討論...
英國計畫從 2018 年開始，超過 250 人的公司必須公佈男女的平均薪資及 Bonus：「Companies will be forced to reveal their gender pay gap」：
The new rules, revealed on Friday, will apply to all companies with more than 250 employees.
除了平均薪資以及 bonus 外，還必須公開每個區間的人數：
In addition to publishing their average gender pay and bonus gap, around 8,000 employers across the country will also have to publish the number of men and women in each pay range.
The government is hoping that naming and shaming firms that pay women a lot less than men in the same jobs will push them to stop the practice, because it will make it harder for them to attract top talent.
In the U.S., similar plans are also under discussions. President Obama announced a proposal earlier this month that would require companies with more than 100 employees to report how much they are paying their employees by race, ethnicity and gender.